Business Object Design: Full Lifecycle Method from Identification to Governance

This article, the third in a series on business objects, details a systematic methodology for designing business objects across their full lifecycle — from identification to governance — so that core business concepts become structured, manageable data assets aligning business, data, and systems.

Core Design Thinking

The approach follows three guiding principles:

Big-picture thinking : Ground object definitions in the enterprise business architecture and organizational structure to ensure semantic consistency and clear boundaries; define business scope, scenarios, and processes to avoid isolated, duplicate, or conflicting definitions.

Small start : Begin with actual process inputs/outputs, information items, and existing IT system data to gradually abstract standard business objects; avoid vague conceptualization and ensure every object can be instantiated and deliver data value.

Closed-loop governance : Treat each business object as the minimum granularity for data governance; bind a responsible department and data owner to every object to enforce "one data one source, one source multiple uses."

Four Phases, Twelve Activities

The methodology unfolds in four sequential phases, each producing concrete deliverables.

Phase 1: Requirements Analysis Preparation

Goal: Build the foundational framework, clarify business scope, scenarios, and data sources.

Analyze business/organizational architecture, define business scope

Map enterprise business capability framework, end-to-end processes, and departmental responsibility boundaries.

Set the analysis boundary for object design to provide a global reference for later abstraction.

Analyze key business scenarios

From the enterprise value chain or government service flow, identify critical user/customer work scenarios at each stage.

Scenario analysis helps pinpoint core data likely to become business objects.

Analyze process activity inputs and outputs

Document input and output information items at every node of end-to-end processes.

Include forms, documents, events, approval data, forming an information-flow baseline table.

Analyze existing IT system models

Inventory current IT systems and database logical models; identify key data objects already managed.

Analyze data operation scope and frequency to inform business object definitions.

Analyze mainstream software package models

Reference data models from SAP, Oracle, Kingdee, etc., to identify standard objects and master data categories.

Avoid local definitions that deviate from industry standards, ensuring extensibility and generality.

Phase Deliverables : Business scope list, core business scenario list, process–information-item mapping table, existing system and software package analysis report.

Phase 2: Identify Business Objects

Core step: abstract from information items to business objects.

Organize candidate business objects (four characteristics)

Pre-screen using four traits: criticality, unique identification, relative independence, instantiability.

Extract data units with independent business meaning from information items, process nodes, and data tables.

Merge and abstract business objects

Unify naming for synonyms (e.g., "customer" vs. "user").

Split homonyms with different meanings (e.g., "order" may separate into sales order and purchase order).

Deduplicate, categorize, and conceptually abstract candidates: merge overly granular objects, standardize fuzzy concepts.

Output business object draft

Draft includes object name, definition, key attributes, information sources, process mapping relationships.

Produce a "business process → information item → business object" mapping table for full traceability.

Phase Deliverables : Business object draft list, object attribute drafts, process–object mapping table.

Phase 3: Validate Business Objects

Systematic validation ensures objects are rational, complete, and integrable.

Rationality validation

Judge whether each object is necessary for business operations, not redundant, and holds independent value.

Optimize based on business importance, complexity, and management responsibility.

Completeness validation

Cross-check against business processes and management needs to ensure all key information is covered.

Prevent missing critical objects or overly narrow definitions.

Integration validation

Verify relationships among business objects and with cross-domain objects.

Ensure design supports system integration and data sharing, avoiding new information silos.

Phase Deliverables : Revised business object list, validation report, object optimization notes.

Phase 4: Finalize Business Objects and Assign Owners

Close the governance loop, turning objects into published enterprise data assets.

Update and publish business object catalog

Refine object definitions, attributes, data sources, lifecycles.

Produce standardized, structured business object assets.

Assign data owners

Bind each business object to a responsible department and individual.

Data owner is accountable for data quality, sharing, and updates.

Formally release the catalog

Enter objects into the enterprise data asset catalog or metadata management platform.

Serve as the unified standard for data governance and system construction, realizing "one data one source, one source multiple uses."

Phase Deliverables : Final business object catalog, data owner assignment table, formally released version.

The article references the book Huawei Business Object as a foundational source and includes a process diagram illustrating the overall flow.